Working Paper: NBER ID: w27600
Authors: Angus Deaton
Abstract: Randomized controlled trials have been used in economics for 50 years, and intensively in economic development for more than 20. There has been a great deal of useful work, but RCTs have no unique advantages or disadvantages over other empirical methods in economics. They do not simplify inference, nor can an RCT establish causality. Many of the difficulties were recognized and explored in economics 30 years ago, but are sometimes forgotten. I review some of the most relevant issues here. The most troubling questions concern ethics, especially when very poor people are experimented on. Finding out what works, even if such a thing is possible, is in itself a deeply inadequate basis for policy
Keywords: randomized controlled trials; economic development; causality; ethics; policy implications
JEL Codes: C01; C93; O22
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| RCTs do not inherently simplify statistical inference (C90) | complexities such as the need for careful data collection and management (C80) |
| Randomization does not automatically balance treatment and control groups (C90) | any observed differences post-treatment cannot be solely attributed to the treatment itself (C90) |
| Self-selection in treatment allocation (C90) | biased results (J15) |
| No RCT can claim to establish causality definitively (C90) | the results may be influenced by confounding factors that are not accounted for in the experimental design (C90) |
| Findings from RCTs in one context cannot be assumed to apply universally (C90) | without considering the specific circumstances of different settings (I00) |
| Ethical concerns regarding RCTs conducted in poor countries (C90) | the results may not reflect the true needs or desires of the subjects involved (C90) |